The class of immunotherapies known as checkpoint inhibitors have proven to be a highly effective and advantageous treatment option for many cancer patients, but they don't work well for everyone, and oncologists have no reliable way to determine in advance which patients are likely to respond to these drugs.
Looking to identify a genetic profile to predict how patients will respond to checkpoint inhibition, also known as anti PD-L1 therapy, researchers collaborated to sequence the tumors of patients who had completed treatment with checkpoint inhibitors. They reported the results of this analysis at the 2017 ASCO Annual Meeting (Abstract 11565).
“While some patients will respond well to checkpoint inhibitors, many will see their tumors progress quickly, so giving them these treatments actually could hurt more than it helps. The task of developing a predictive tool to help oncologists determine which patients will and will not respond to anti-PD-L1 therapy is a matter of high priority,” stated Carl Morrison, MD, DVM, Executive Director of the Center for Personalized Medicine at Roswell Park, Buffalo, N.Y.
The research team discovered a set of 54 immune-related genes using expression data from whole transcriptome RNA-sequencing in 300 tumor specimens. They then designed a sequencing test to measure both gene expression and tumor mutational burden from a population of 167 Roswell Park patients previously treated with approved checkpoint inhibitors, with complete treatment and response data available for 87 patients. From these measurements and data, the team created and tested an algorithm designed to predict clinical response to checkpoint inhibition. Their algorithm, which incorporated expression data for those 54 genes along with a patient's mutational burden, accurately reflected therapeutic response for 90 percent of patients. By comparison, an analysis based on positive PD-L1 immunohistochemistry results or high mutational burden status alone correctly predicted response in only 30 percent of cases.
“Our results show that this two-pronged algorithm—expression analysis for the 54-gene signature we identified, on top of mutational burden—may prove to be a highly accurate tool for predicting treatment response,” concluded Morrison. “We're excited by these striking findings and look forward to validating our results in a larger study.”